env argv0 overwrite possibility (unix only) - fixes new gnu test version

[cre4ture]

as discussed with @sylvestre here: #5801

this makes the gnu-test tests/env/env.sh green again.

changes explained:

• accept new parameter and provide it at program start
• adapt debug logmessages to fit to gnu ones
• introduce a second debug-level (-vv) that allowes more detailed, non-gnu conform logs.
• add own test that uses --argv0 argument to coreutils executable to execute specific tools without placing seperate argument for tool-name
• UPDATE: patch gnu test to seperately test "-a" and "--argv0"

[github-actions]

GNU testsuite comparison:

Skipping an intermittent issue tests/tail/inotify-dir-recreate (passes in this run but fails in the 'main' branch)

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/dd/stats is no longer failing!

[sylvestre]

GNU testsuite comparison:

Congrats! The gnu test tests/dd/stats is no longer failing!

are you sure you fixed it ? :)

[cre4ture]

are you sure you fixed it ? :)

sorry, I was apparently a bit quick there. I forgot to add the patch for the test.
And here I need your opinion.
The GNU-test is kind of crappy due to two reasons:

1. it prepares the expected output but doesn't actually compare it with the real output.
2. it tests both option-variants at the same time. It provides the short form -a and the long form --argv0 at the same time expecting that the last one provided precedes. uutils/clap by default raises an error if one is doing this, and I think this makes actually more sence. Thats why I provided the patch.

What do you think. Is it more important to replicate the GNU behaviour in this case?

[sylvestre]

We could report bugs upstream to get this adjusted.

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/tail/inotify-dir-recreate (fails in this run but passes in the 'main' branch)

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!

[BenWiederhake]

I might be missing some context, but:

it tests both option-variants at the same time. It provides the short form -a and the long form --argv0 at the same time expecting that the last one provided precedes.

That's actually the case for the vast majority of options (but not all of them, sigh), and there's a huge chunk of programs where we still do it wrong: #5998

In this case, simply doing Arg::new("argv0").overrides_with("argv0") should do the trick.

[cre4ture]

@BenWiederhake thanks for pointing this out.
As our reference is gnu, I will change it even though I do still think that our initial implementation (self-conflict) is more intuitive. Why would someone intentionally provide twice the same option?

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!
Skipping an intermittent issue tests/tail/inotify-dir-recreate (passes in this run but fails in the 'main' branch)

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!
Skipping an intermittent issue tests/tail/inotify-dir-recreate (passes in this run but fails in the 'main' branch)

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!

[github-actions]

GNU testsuite comparison:

Skip an intermittent issue tests/rm/rm1 (fails in this run but passes in the 'main' branch)
Congrats! The gnu test tests/env/env is no longer failing!
Skipping an intermittent issue tests/tail/inotify-dir-recreate (passes in this run but fails in the 'main' branch)

[BenWiederhake]

Looks mostly good, just a few nitpicks about testing.

[BenWiederhake]

Fantastic! I didn't realize that argument overrides across -S already works.

I'm a bit uncertain how to add this to the repo; squashing it all seems like a bad idea (e.g. the changes to GNUmakefile should be in a different commit than the actual argv0 changes), but the commits are also not atomic, so I can't just rebase it. I'll discuss this with the others.

In the meantime, please consider doing git-rebase -i origin/main :)

[github-actions]

GNU testsuite comparison:

GNU test failed: tests/timeout/timeout. tests/timeout/timeout is passing on 'main'. Maybe you have to rebase?
Congrats! The gnu test tests/env/env is no longer failing!

[github-actions]

GNU testsuite comparison:

Congrats! The gnu test tests/env/env is no longer failing!

[BenWiederhake]

Awesome work, and thanks for the rebase! :)